Yes, I actually bought a second DX just to see if I had a faulty one. The issue boils down to heat creep/retraction. The device wants to retract too much and wears down the filament causing it to stay in place and clog.
Works fine as a single nozzle device. I’d recommend developing something similar to a AMS/MMU from Bambu Labs and dumping the dual nozzle.
Hello everyone,
I have found another source for Bondtech MK8 Heat Break!
Hello everyone,
As Snapmaker has now released a Hotend Kit Update, I would like to draw your attention to it.
Of course, it will take a while until more accurate results of daily use will be available.
Please post the results of the new Hotend-Kit in the section mentioned.
Now I wanted to buy some spare heat breaks - but those guys do not ship to outside the UK 
We have updated the clogging issues and new hotend kit. Check it out:
In case someone does not like the new hotend due to it not being able to be taken apart by the user (i.e. if the nozzle is broken, it very much seems you have to buy a complete new module!), I have found another european source for the Copperhead MK8 heat break - this one is in Spain:
In the US or Canada, there are also some suppliers that still seem to have it:
and
Please keep in mind that I do not know those suppliers and have only found them by searching. Therefore check if they are trustworthy, as you should do with every online shop 
The MK8 version seems to have been discontinued by Slice Engineering. Not sure what replaces it.
Discontinued notice: https://www.bondtech.se/product/copperhead-mk8-heat-break/
Other options: Copperhead® Bimetallic Heat Break | Slice Engineering
Yes, this is known and indeed unfortunate… and the reason for the posts above that link to sellers that still offer them. We are bound to the MK8-like connection with M6 thread that the J1 heat break has.
The only alternative I see so far is the posts from @Viktor above about that Aliexpress MK8 bimetal heat break.
This one might be an alternative as well, it is closer to the original Copperhead design:
Has anyone tried these Mk8 heat brakes in light of the discontinuation of the slice engineering copperhead version?
those are exactly what I’m using along with volcano adapter for spacers and so far no problems whatsoever
As soon as new hardened hotends became available from Snapmaker, I decided to have them and give a try. They worked great indeed for me in different materials, also silky PLA. As they are press fit with no chance of nozzle change as separate unit, I decided to keep them for times I might need to print some abrasive stuff.
It has happened so for the time being I mostly run my SJ1 for parts made of cheapest available PLA, fancy hotends for this is large overkill. So I decided to remove precious new ones and switch back to the ones from “clog fighting times”.
I made the following setup. Left extruder - stock heatsink and fan, Copperhead heatbreak, plated copper heaterblock from E3D and AP3X nozzle by SliceEngineering. Mostly I did this because I liked to test GammaMaster AP3X nozzle and use of Copperhead heatbreak allowed for longer thread of nozzle. On right side I put stock heatsink, fan, heatbreak (!) and heaterblock, just MK8 type plated nozzle from MicroSwiss. Just assembly this time was done with proper amount of heat paste.
Now for this setup more than 600 h total printing time is over, prints ranging from 2 to 15 hours each, 24/7 based printing as much as possible, all the time in Copy mode with both extruders involved. Printing with doors closed, lid off (I have no additional fans anywhere in SJ1). Believe or not - I had zero clogs during over a month of almost uninterrupted printing. Temperature insides the printer stays around 35 C. I use the same PLA I had to work with when heat creeping newly purchased SJ1 was ruining my mental comfort and I had to join Copperhead heatbreak community on this forum.
So results of this rather intensive testing have hardened my belief the only basic problem with stock extruders on newly launched SJ1 was lack of heat paste during assembly of hotends. I had 6 of them with printer when purchased it and it was hard to find traces of heat conductive paste in any of them.
I also shortened one of original titanium heatbreaks to accomodate GammaMaster AP3X 0.6mm nozzle for CF filaments and it works flawless too. Of course, this is not PLA any more.
I beg to differ with this paragraph
My guess is you did update the printer firmware and/or Luban at some point in between your tests, correct?
Unless my assumption is wrong, the thing you did prove with this test is that a sufficient amount of thermal paste combined with the lowered retraction limits Snapmaker introduced at some point solves the problem, at least for your material(s).
Since the amount of compliants as reduced significantly since the “clogging fight”, this is something I did expect to see 
But anyway: the important thing is that the J1 runs better now than it did in the beginning. This is what counts.
Sure, I completely agree to your statement of significant positive impact of lowered retraction limits!
I did firmware update on my SJ1 once and also Luban has been updated on more or less regular basis. But I remember when I had to face “clogging fight”, I lowered retraction setting in Luban to 1mm right away (it was some higher in the beginning) and since that has been checking it from time to time to avoid surprises. To tell the truth, I have not been following what improvements firmware or Luban updates bring so probably there has been positive impact I had no clue about indeed.
Anyway - J1 has become more reliable over the year of turbulence
Funny to admit on firmware updates. I did mine just because I wanted to try PID tuning feature with new hardened hotends. It worked well. But this PID tuning feature does not work (it is not available in menu, at least I didn’t find it) when regular hotends, also custom modified ones, are used.
I cannot confirm what the Snapmaker team did with Luban since I never used that software, but they did change at least two things in their Cura plugin:
If I remember our tests correctly, that last one had a significant positive impact on clogging - but it also had unwanted side effects in some cases (very small structures with lots of retracts) since it simply blocked any retracts if that amount was reached.
I very much guess they implemented something similar in Luban.
@Karass3D @Mechanikus Yes. We have updated the retraction settings since Luban V4.7.3 and Snapmaker Cura Plugin V0.9.0, which helped to avoid some clogging issues caused by inappropriate retraction settings. Learn more in this Wiki: Inappropriate print settings cause a clog | Snapmaker Wiki
Hello, as hundreds of users have already received the new hotends, we are interested in collecting feedback on the J1/J1s new hotends from these users. We’ve sent out an email to users who have received the new hotends in mid-May, but some recipients may have missed it. To ensure we gather ample feedback from users who have utilized the new hotends for a while, we are sharing the survey link here. Your input is highly valuable to us, and we truly appreciate your sincere feedback. Kindly click on the link Product Feedback - J1/J1s New Hotend to fill out the survey and share your experience with the new hotends .
If you have already submitted the survey, please disregard this message.
Welcome to participate in the discussion by following this thread:
I have not had a single clog since getting the new hot ends 
Hi,
many people mentioned there was thermal paste missing on the old hotends. Can anyone share a photo of where exactly the paste is missing, and how to get there? Can you use the paste that usually comes with CPUs?
Look here: Potential fix for J1 clicking and clogging - #3 by nnowak
When you unscrew the heat break from the heat sink, there is a good chance you just see a blank M6 thread just with some minor traces of thermal paste if you are lucky. The thread should be covered in thermal paste to provide a decent heat conduction though.
From my point of view it is however easier to unscrew the fan and PCB from the heat sink instead of unscrewing the temperature sensor and the heater cartridge from the heat block as nnowak did there. But that is personal preference. Just take some photos andd make sure the wires will run in the same way they did prior to disassembly when you screw everything together again.
Regarding thermal paste: yes, a good CPU thermal paste does the job here as well since it is the cold end of the heat break that gets the stuff. I used Arctic MX-4.
Never posted it here. The combo I used was:
I recall it was tricky getting the orientation of the heat block correct while also getting the right depth into the block.
I also recall I didn’t tighten/heat the block properly. Came back to find the left nozzle was oozing from the top once while printing, but caught it in time to not be a big, big mess.
Also the reminder to check the electrical continuity to Pin 1 of the PCB to the nozzle tip before installing. That was good advice. One of them didn’t the first time I assembled it. Start over…
Anyways. These nozzles eliminated frequent, minor skips and clogs at any random layer while printing PLA when many other suggested fixes didn’t. Every once in a while, I instead found I would sometimes get a hard clog during the first layer of a print.
This was new. I can’t confirm this solved it, but I then started preheating the bed. Theory being it was due to the bed and hotend starting at the same time by default start gcode, but the bed takes much longer.
(Also later changed the start gcode to wait for bed before firing the hotend in case I forget. Usually I remotely heat the bed while I’m slicing. )
Now I think I haven’t really seen this first layer clog happen again.
I did not do any flow testing since I don’t really think the kinematics are good above 175 mm/s without getting a shift every once in a rare while. The Y gantry isn’t light.
Lastly, the updated Snapmaker, all-metal hotends came out a few months later and I switched to those. These were provided freely to me as an affected person when they did the community outreach about it.
In terms of performance (other than flow which I’ve not tested), they’ve performed identically to the ones I made as far as I can tell. Including a first layer clog when I didn’t preheat the bed, and no skips while printing. So I’ve just kept using those since they’re also hardened.
Cheers!